M5 Improving Chatbot client error


Although all my graded cells are running succesfully without errors when I am submitting to the coursera grader it shows client not found.I also tried importing client and assigning separately for my graded cells but still it is showing same.How can I fix this? Someone please help.

did you try once clearing out kernel output, reconnect and re-run all the cells and submit??

Everything.

can you DM me the screenshot of the codes for exercise you failed. click on my name and then message.

It shows your public profile is hidden.Also there are no exercises I failed.All unit tests are running successfully. It is when I am submitting it to the grader, it shows client not declared which is part of the ungraded component. I think I should take it upto Coursera support for fix or is there anyway round this?

I am sending you msg, in case you want to send codes to review.

passing all unittest doesn’t always confirm your correct as per autograder.

like if a user as changed or edited too many parts of the assignment notebook, grader fails. if user has used grade-up-to-here, can also cause grade failure and lastly if there are some hard-coding done, than what autograder looking for, then also grades fail, hence I told to send your codes screenshot for review.

Likely that is a bad idea, and probably made the situation worse and more confusing.

Another cause of grader problems is if you renamed any of your assignment files.

@Sujato

The mistake in your codes is only that in both exercises you are separately recalling the client response for the faq collection and product collection which isn’t required in both exercises you are failing.

just using query.near_text with query and limit argument to both collections was enough to get your exercise to clear.

Note that for exercise 4 your other written codes are perfectly alright.

Thank you it worked. Although the last cell isn’t working but I got 75%, enough for passing the assignment.Really appreciate your help.

hi @Sujato

i would recommend to clear all the grading task as you are in a learning journey.

I agree getting 100% isn’t a must criteria here but just passing also shouldn’t be the comfort as we tend fail. interviews, exams by just few marks. Getting better at understanding and writing and debugging codes should be the atmost aim.

You never know the exercise you didn’t pass this assignment might teach you something worth learning! about python coding, ai, or whatever you aim to do.

Otherwise I leave it to you.

Good luck, keep learning! try debugging the exercise once again after taking break. You can send me codes for review if you are not able to understand.

Regards

DP